Anda di halaman 1dari 8

J. theor. Biol.

(2002) 215, 441–448


doi:10.1006/jtbi.2001.2523, available online at http://www.idealibrary.com on

Species Positions and Extinction Dynamics in Simple Food Webs


Ferenc JordaŁ n*wz, IstvaŁ n Scheuringy and GaŁ bor Vidaw

n
!
Collegium Budapest, Institute for Advanced Study, Budapest H-1014, Szentharoms ! u. 2., Hungary
ag
wDepartment of Genetics, Evolutionary Genetics Research Group, Eotv
. os
. University, Budapest H-1117,
! any
Pazm ! P. s. 1/c, Hungary and y Department of Plant Taxonomy and Ecology, Research Group of
. os
Ecology and Theoretical Biology, Eotv ! any
. University, Budapest H-1117, Pazm ! P. s. 1/c, Hungary

(Received on 22 June 2001, Accepted in revised form on 18 December 2001)

Recent investigations on the structure of complex networks have provided interesting results
for ecologists. Being inspired by these studies, we analyse a well-defined set of small model
food webs. The extinction probability caused by internal Lotka–Volterra dynamics is
compared to the position of species. Simulations have revealed that some global properties of
these food webs (e.g. the homogeneity of connectedness) and the positions of species therein
(e.g. interaction pattern) make them prone to modelled biotic extinction caused by
population dynamical effects. We found that: (a) homogeneity in the connectedness structure
increases the probability of extinction events; (b) in addition to the number of interactions,
their orientations also influence the future of species in a web. Since species in characteristic
network positions are prone to extinction, results could also be interpreted as describing the
properties of preferred states of food webs during community assembly. Our results may
contribute to understanding the intimate relationship between pattern and process in
ecology.
r 2002 Elsevier Science Ltd. All rights reserved.

Introduction indicating robust functional insensitivity against


Interesting ideas in ecology frequently come random errors (Jeong et al., 2000).
from distant fields of science. The conclusions of The dynamical behaviour of a well-defined set
recent studies on the structure of complex of small sink webs (Cohen, 1978) has been
networks (Watts & Strogatz, 1998; Baraba! si & studied (Jorda! n & Molna! r, 1999; Jorda! n et al.,
Albert, 1999; Albert et al., 2000; Jeong et al., submitted), giving rise to the problem of how the
2000; Amaral et al., 2000; Williams et al., network position of frequently extinct species
submitted), giving information about how meta- can be characterized. Since the large-scale
bolic and neuronal networks seem to be orga- structure of biological networks depends highly
nized, may also be useful for ecologists. An on internal dynamical processes and constraints
interesting result is that metabolic networks (Sugihara, 1984), we analyse how the internal
follow a power-law connectivity distribution extinction dynamics (based on Lotka–Volterra
models) in these webs is related to the topology
of trophic interactions. We think that these basic
zAuthor to whom correspondence should be addressed. equations, even if their applicability is limited,
E-mail: jordanf@falco.elte.hu are good as a starting point for further analyses.

0022-5193/02/$35.00/0 r 2002 Elsevier Science Ltd. All rights reserved.


442 F. JORDÁN ET AL.

Methods The orientation of neighbours is characterized


by D0 ¼ DinDout (where Din is the in-degree: the
We compared the extinction dynamics of number of links directed to a particular node,
species constituting 25 model sink webs (i.e. food and Dout is the out-degree: the number of links
webs with a single top-predator, see Fig. 1). This coming from a particular node). Low and high
set of webs contains all of the 25 topologically values of D0 characterize multiple exploited
different model sink food webs possibly built up producers and generalist consumers, respec-
from five species (graph nodes) and five trophic tively, in the directed graph.
interactions (links). Since the connectedness of The concept of ‘‘net status’’ (Harary, 1959,
these webs equals uniformly C ¼ 0.5, we are able 1961) has been modified slightly in order to be
to analyse the effect of link pattern on population ecologically more reasonable (see Jorda! n et al.,
dynamics, distinguished from the effects of 1999; Jorda! n, 2000, 2001). The modified index
changing connectedness. Food webs and indivi- was constructed to give a quantitative measure
dual species are characterized by the following of the importance of species (species of
structural and dynamical properties. outstanding importance are called ‘‘keystone
The degree (D) of a graph node is the number species’’, Paine, 1969; Power et al., 1996). This
of its neighbours (both prey and predators in the is the positional keystone index (K ) which
food web graph). Because all these webs contain considers not only neighbours in the web (like
five links, the sum of D values (sumD) equals D) but their neighbours as well, thus providing
10 for every web. However, the D values of more information about network structure. If
individual species differ; this difference is mea- bottom-up and top-down forces are considered
sured by the standard deviation of D within a to be of equal importance, then let K ðiÞ ¼ KbðiÞ þ
web (stdD). KtðiÞ ; where KbðiÞ and KtðiÞ refer to the bottom-up

1 1 2 1 3 1 4 1 5 1 6 1

2 2 2 2 2 3 2
3 4 5 3 4 5 3 4 5 3 3
4 5 4 5 4 5
7 1 8 1 9 1 10 1
11 1 12 1 13 1
2 2 2
3 2 3 2
2 2
3 3
3 3 3
4 5 4 5 4 5
4 5 4 5 4 5 4 5
1 17 1
14 1 15 1 16 18 1 19 1 20 1
2
2 2 2 2 2
2
3 3 4
3 3 3 3 3
4 5 4 5 5 4 5 4
4 4
5 1 1 5 5
23 24 1
1 25
1 22
2 2 2
21 2 3 3
2 3 3
4
3 4 4 4 4
5 5
5 5 5

Fig. 1. Twenty-five model sink webs containing five species (nodes) and five trophic links. Both species and webs are
numbered for further reference. Higher species consume lower ones (species #1 is always the top-predator; the direction of
links is not shown for simplicity). For data on structural (D, D0 , K) and dynamical (E) characterization of species, consult
http://falco.elte.hu/Bjordanf/data2.prn.
EXTINCTION DYNAMICS IN FOOD WEBS 443

and top-down effects. These indices count the j on the per capita reproduction rate of species
number of disconnected species after the re- i. Here, ai is a uniformly distributed random
moval of species i. If species i is in an important variable in the range (0 1) if species i is a basal
network position it has a rich system of trophic species, while ai is selected randomly from (1 0)
interactions, then deletion will lead to more if species i is intermediate or top species.
disconnections in both directions. The effects of Similarly, bij falls into (0 1) if i is the pre-
species deletion on the disconnection of both dator of j, and bji is selected from (1 0) if i
bottom-up material flows and top-down flows of eats j. If i ¼ j, bij is selected from (1 0) in case
trophic regulatory effects are quantified by these of producers but equals zero for higher species
indices of positional importance. The Kb value of (May, 1973; Pimm, 1982). The dynamics of
the i-th species can be calculated as system (2) is simulated with 2000 random
n   parameter sets selected in the way described
X 1 1 ðcÞ
ðiÞ
Kb ¼ þ K ; ð1Þ above. We considered a species to be extinct
c¼1
dc dc b if its density fell below a critical level (here,
d ¼ 1015; results do not differ qualitatively
where n is the number of its predators, dc is the for other values of d) after a given time (here,
number of prey eaten by its c-th predator, and t ¼ 200 steps). The extinction value (E) of a
KbðcÞ is the bottom-up keystone index of its c-th species gives the number of extinctions out of
predator. Thus, the bottom-up index should be 2000 simulations (we suggest that our approach
calculated first for top species. Kt is calculated to stability is more general than to study stable
similarly after turning the web upside down. It fix points).
should be emphasized that K refers to keystone
species only in a trophic sense (i.e. importance in
maintaining trophic network flows). K charac- Results
terizes food web pattern at an intermediate scale, The distribution of the basic structural indices
between the local (measured by D and D0 ) and (D and K) of species in these model webs is
global (measured by stdD) properties of webs. shown in Fig. 2: the majority of species have two
Nevertheless, in our small webs K seems to be neighbours (D ¼ 2), while K is distributed evenly
a rather global index. Both the sum and the (especially if sink species are not considered).
standard deviation of K-values within webs The most frequently extinct nodes can be
(sumK and stdK) were also calculated for characterized by an intermediate degree (D ¼ 2,
analyses. see Fig. 3). The keystone index itself does not
We are interested in whether these structural seem to be in correlation with the probability of
indices are possibly as informative and interest- extinction (Fig. 3; but seemingly extinctions are
ing as traditional food web statistics having been not frequent with very low K). However, the sum
used extensively in the literature (e.g. trophic of extinction events within a web is clearly higher
height, compartmentalization, the level of if the sum of keystone indices is higher (Fig. 3).
omnivory, etc., see Pimm, 1980). The fact that K correlates with E only at the level
The behaviour of species is characterized by of whole webs reflects how keystone index is
dynamical Lotka–Volterra models. Population calculated: it also quantifies indirect effects,
dynamics is described by the following set of which are of more global nature than locally
equations: understandable direct interactions. In order to
! study the effects of both D and K on extinction
Xn
x’ i ¼ ai þ bij xj xi ; i ¼ 1; y; n: ð2Þ dynamics, they are plotted together (Fig. 4). If
j¼1 D ¼ 1, species go extinct with high probability if
K ¼ 4; however, the probability of extinction is
Here, xi is the population density of species i, low at smaller K indices. If D ¼ 2, extinction can
ai is the per capita rate of reproduction (for be very probable only with low K values. In case
growth, ai40, for decay, aio0). The parameter of a ‘‘dangerous’’ number of neighbours (i.e. if
bij indicates the per capita effect of species D ¼ 2), additional indirect effects (e.g. K ¼ 4) may
444 F. JORDÁN ET AL.

Fig. 2. The frequency distribution for the degree of nodes is a discrete, unimodal distribution (on the left). K values are
evenly distributed (on the right; note that for all nodes of the last column K ¼ 4: 25 out of 47 nodes represent the sink species;
thus, basal and intermediate species are distributed much more evenly).

Fig. 3. Relationships between structural and dynamical properties: top left, the number of extinction events concerning
each species (E, out of 2000 simulations) is plotted against the number of their neighbours (D, ( ) show the average; the
Mann–Whitney test on a distribution of 105 randomly generated samples shows significance at po0.0003, po0.0114, and
po0.0025 for differences between [D ¼ 1, 2], [D ¼ 2, 3], and [D ¼ 3, 4] pairs of categories, respectively); top right, the number
of extinction events (E) is plotted against species’ keystone indices (K); bottom, the sum of extinction events (sumE) within
each web is plotted against the sum of their species’ keystone indices (sumK ).

decrease the probability of dynamical extinction. exclusively direct (D) and direct plus indirect (K)
If D ¼ 3 or 4, extinction is less probable and effects gives an interesting extinction probability
is less sensitive to K. Thus, measuring both landscape.
EXTINCTION DYNAMICS IN FOOD WEBS 445

Fig. 4. The effects of the keystone index (K) on extinction dynamics (E), for nodes characterized by different D values
(D ¼ 1, 2, 3, and 4). If a species has a single direct interaction (D ¼ 1), then it is more dangerous to have stronger indirect
connections (large K). However, in case of having two neighbours (D ¼ 2), species can go extinct more frequently only if K is
lower. For D ¼ 3 and 4, extinction frequency depends weakly on K is lower. For D=3 and 4, extinction frequency depends
weakly on K.

Since both the number of neighbours and their measure how homogeneously the species are
orientation can be important for extinction connected. High stdD means, for example, one
dynamics, we analysed the relationship between strongly and ‘‘many’’ poorly interconnected
D0 and E (Fig. 5). For example, if D ¼ 2, D0 may species, while a low stdD indicates that all
equal 2 (two prey of a sink species, e.g. Fig. 1, species have roughly equal connectedness. Simi-
web #3, species #1), 0 (one prey and one larly, stdK quantifies the homogeneity of direct
consumer of an intermediate species, e.g. and indirect connectedness. Homogeneous webs
Fig. 1, web #19, species #3) and 2 (two can be characterised by a slightly higher sum of
consumers of a source species, e.g. Fig. 1, web extinctions (Fig. 6). This finding may imply that
#21, species #5). The extinction probability (E) is species in key positions can, in general, increase
the highest for species with D0 ¼ 0. Thus, both dynamical stability at the community level.
multiple exploited producers (D0 ¼ 2) and
generalist consumers (D41) are at lower risk
Discussion
of extinction. ‘‘Generalist’’ feeding and ‘‘multi-
ple’’ food supply is dynamically more advanta- Because the pattern of population dynamics is
geous, even if the number of direct interactions is not uniform for the species of our model webs,
the same. positional effects are clearly demonstrated (i.e.
Within a web, the standard deviation of quantified position within the network gives
degrees (stdD) and keystone indices (stdK) information on the dynamical properties of
446 F. JORDÁN ET AL.

Fig. 5. The number of species extinction events (E) is plotted against species’ D0 indices ( ) show the average; see
explanation in text): not only the number but also the orientation of direct trophic links influence extinction dynamics. The
Mann–Whitney test on a distribution of 105 randomly generated samples shows significant difference between [D0 ¼ 1, 0] at
po0.00001. [D0 ¼ 0, 1] and [D0 ¼ 1, 2] show different average E at po 0.1. D0 ¼ 2 and 3 differ significantly at po0.003, while
[D0 ¼ 2, 1] and [D0 ¼ 3, 4] do not differ significantly.

Fig. 6. The sum of extinction events (sumE) for each web is plotted against the standard deviation of degrees (stdD,
( )) and the standard deviation of keystone indices (stdK, ( )) of species. Living in a homogeneous web is more
dangerous.

nodes). These positional properties of nodes may deletions in large networks, see Albert et al.,
characterize either extinction risk (as far as 2000). For this set of small sink webs, it can
randomly chosen interaction parameters are be concluded that the presence of species in key
regarded as perturbations) or preferences in positions can be dynamically either advanta-
community assembly (as far as possible struc- geous [if only direct interactions are considered
tures created by invaders are compared). If (see stdD in Fig. 6)] or disadvantageous [if
positional properties (D, D0 , K) influence extinc- both direct and indirect interactions are
tion dynamics (E), species do not die out taken into account (see Fig. 3)] against these
randomly, but rather in an internally directed ‘‘internal attacks’’ [in contrast with Albert et al.
way (alike ‘‘attacks’’ are defined as directed node (2000), where non-random deletions are more
EXTINCTION DYNAMICS IN FOOD WEBS 447

dangerous in heterogeneous networks]. The scale, (1) species have either a single or ‘‘many’’
description of extinction-prone network posi- (3 or 4) direct trophic links (see, for example,
tions is suggested to identify unpreferred struc- species #5 and #1 of web #1, Fig. 1), (2) the
tural arrangements during web assembly neighbours are clustered in either bottom-up or
(cf. Sugihara, 1984): the success of invasion top-down direction (D0 a0, see species #2 in web
may depend characteristically on the particular #3); at an intermediate scale, (3) they have either
position of the invader within the community a single neighbour and a low K-index or two
network (and this may hold also in case of neighbours and a high K-index (see species #5 in
speciation). web #7 and species #5 in web #21), (4) the sum
The main mechanisms behind the simulated of K-indices within a web is low (see web #3),
network dynamics include direct (e.g. predation, and, at the global scale, (5) whole webs have
food supply) and indirect (e.g. exploitative differently connected nodes (heterogeneity, stdD,
competition, trophic cascade) trophic interac- see web #2). We have to emphasize that the
tions among species (see a catalogue of indirect simplest dynamical model was analysed: asym-
interactions in Menge, 1995). Given the topology metries in interaction strength (bij parameters)
of a trophic network, one may predict either the may lead to qualitative differences in our
nature of a particular interaction or the fate of a conclusions (Jorda! n et al., submitted).
particular species, but the dynamics of whole If trophic control in a particular community is
webs remain unclear until simulated. Only joint considered of primary importance, i.e. food web
dynamical equations can give higher-level pre- studies have strong predictive power, then
dictions (e.g. the fate of a whole web). Our linking trophic structure to species’ dynamics
simulation results show typical patterns in the could provide useful information for conserva-
dynamics of food webs and their species. For tionists. The quest for keystone species (Power
example, let us consider food web #12 in Fig. 1. et al., 1996) is a big challenge for community
This web can be considered typical as for the ecologists. Quantitative approaches to identify-
frequency of extinctions (1760 extinction events ing keystone species in ecosystems (e.g. based on
occurred out of 2000 simulations). The extinc- food web position) could increase the predictive
tion frequency of the top-predator (#1, E ¼ 482) power of studies on the relative importance of
is close to the average. Its position is neither too species within communities. In addition, if key-
risky nor very safe. One of its prey (#2, E ¼ 832) stones really make communities less prone to
is prone to extinction, as are species with D0 ¼ 0 extinction, we can get closer to explaining their
in general. They have a single source to consume outstanding importance. Thus, the prediction
and their fate depends strongly on their pre- of extinction dynamics by food web analysis
dators’ behaviour. Its other prey (#4, E ¼ 241) could give a useful tool for conservationists
has a high extinction probability compared to (cf. Simberloff, 1998). If biological diversity can
other producers. It is caused by the following: if be expressed in terms of positional diversity of
the top-predator’s density increases, it exploits species in trophic networks, then the diverse
species #4 both directly (by predation) and population dynamics of species can also be
indirectly (by having a negative effect on species approached quantitatively.
#2, which is good for species #3, which is bad We focused only on a particular but well-
again for species #4). Species #3 (E ¼ 87) is well defined set of small food webs. Thus, possible
saved by topology: it has a single predator and effects of scale- and context-dependence were
two preys. Finally, there is a producer in this neglected. First, it should be clarified whether
‘‘community’’ (#5, E ¼ 118) which goes to our results based on the analysis of small webs
extinction less frequently than the other one would also be applicable for large networks (see
(#4), because its position is better as for the Bersier & Sugihara, 1997). Second, the problem
number of predators. of how the dynamics of community webs differs
We can conclude that dynamically caused from that of sink webs is to be studied. The
species extinctions are less frequent if the behaviour of persistent nodes following pertur-
following conditions are satisfied: at the local bations is to be analysed in detail (some
448 F. JORDÁN ET AL.

preliminary results on whole webs are given in Cohen, J. E. (1978). Food Webs and Niche Space.
Jorda! n et al., submitted). Further, the manner in Princeton, NJ: Princeton University Press.
Harary, F. (1959). Status and contrastatus. Sociometry 22,
which certain network properties affect basic 23–43.
organizational principles, similar to the ‘‘ageing’’ Harary, F. (1961). Who eats whom? Gen. Syst. 6, 41–44.
and ‘‘cost’’ of nodes in the study of Amaral et al. Jeong, H., Tombor, B., Albert, R., Oltvai, Z. N. &
BarabaŁ si, A.-L. (2000). The large-scale organization of
(2000), should be considered. Nevertheless, we metabolic networks. Nature 407, 651–654.
think that this study can contribute to the JordaŁ n, F. (2000). Is the role of trophic control larger in
understanding of the ‘‘pattern and process’’ a stressed ecosystem? Commun. Ecol. 1, 139–146.
JordaŁ n, F. (2001). Trophic fields. Community Ecol. 2,
problems in ecology. Future studies on the 181–185.
homogeneity of ecological networks may reveal JordaŁ n, F. & MolnaŁ r, I. (1999). Reliable flows and
important aspects of the organization of food preferred patterns in food webs. Evol. Ecol. Res. 1,
webs, and, in general, the ecological principles of 591–609.
JordaŁ n, F., TakaŁcs-saŁ nta, A. & MolnaŁ r, I. (1999).
community assembly and biotic extinctions. A reliability theoretical quest for keystones. Oikos 86,
453–462.
We are grateful to Ja! nos Podani for his help in JordaŁ n, F., Scheuring, I. & MolnaŁ r, I. Persistence, flow
statistics and correcting English. F. J. also thanks reliability in simple food webs. Submitted.
. Szathma! ry for enabling a fruitful stay
Professor Eors May, R. M. (1973). Stability and Complexity in Model
Ecosystems. Princeton, NJ: Princeton University Press.
at Collegium Budapest, Mauro Santos and Elias Menge, B. A. (1995). Indirect effects in marine rocky
Zintzaras for some methodical help. Two anonymous intertidal interaction webs: patterns and importance.
reviewers are kindly acknowledged for helpful com- Ecol. Monogr. 65, 21–74.
ments on the manuscript. F. J. is supported by two Paine, R. T. (1969). A note on trophic complexity and
grants of the Hungarian Scientific Research Fund, community stability. Am. Nat. 103, 91–93.
OTKA F 029800 and OTKA F 035092. I. Sch. is Pimm, S. L. (1980). Properties of food webs. Ecology 61,
supported by the grant OTKA T 029789. Both F. J. 219–225.
and I. Sch. are recipients of the Bolyai Research Pimm, S. L. (1982). Food Webs. London: Chapman & Hall.
Grant of the Hungarian Academy of Sciences. Power, M. E., Tilman, D., Estes, J. A., Menge, B. A.,
Bond, W. J., Mills, L. S., Daily, G., Castilla, J. C.,
Lubchenco, J. & Paine, R. T. (1996). Challenges in the
quest for keystones. BioScience 46, 609–620.
REFERENCES Simberloff, D. (1998). Flagships, umbrellas, and
keystones: is single-species management pass!e in the
Albert, R., Jeong, H. & BarabaŁ si, A.-L. (2000). Error landscape era? Biol. Conserv. 83, 247–257.
and attack tolerance of complex networks. Nature 406, Sugihara, G. (1984). Graph theory, homology and food
378–381. webs. Proc. Symp. Appl. Math. 30, 83–101.
Amaral, L. A. N., Scala, A., BartheŁ leŁ my, E. & Watts, D. J. & Strogatz, S. H. (1998). Collective
Stanley, H. E. (2000). Classes of small-world networks. dynamics of ‘small-world’ networks. Nature 393,
Proc. Natl Acad. Sci. U.S.A. 97, 11 149–11 152. 440–442.
BarabaŁ si, A.-L. & Albert, R. (1999). Emergence of Williams, R. J., Martinez, N. D., Berlow, E. L.,
scaling in random networks. Science 286, 509–512. Dunne, J. A. & Barab Asi, ! A.-L. Two degrees of
Bersier, L. F., & Sugihara, G. (1997). Scaling regions for separation in complex food webs. Proc. Natl. Acad. Sci.
food web properties. Proc. Natl Acad. Sci. U.S.A. 94, U.S.A. (In press).
1247–1251.

Anda mungkin juga menyukai